Low prevalence of antigliadin and anti-endomysium antibodies in subclinical/silent celiac disease

OBJECTIVE: Endomysial antibodies (EMA) are a well-known hallmark of celiac disease, but some recent studies showed that the prevalence of these antibo dies in clinical practice is lower than expected. The aim of our study was to determine the prevalence of antigliadin (AGA) and EMA antibodies on a co nsecutive series of subclinical/silent celiac patients. METHODS: We studied 115 consecutive patients with subclinical (92 patients) or silent (23 patients) forms of celiac disease. AGA and EMA were screened in all patients. Histopathology of celiac disease was expressed according to the Marsh classification. RESULTS: The overall AGA in subclinical form were positive in 77% (14 of 18 ) of patients with partial villous atrophy (VA), in 84% (21 of 25) of patie nts with subtotal VA, and in 90% (27 of 30) of patients with total VA, wher eas EMA were positive in 88.88% (16 of 18) of patients with partial VA, in 92% (23 of 25) of patients with subtotal VA, and 96.66% (29 of 30) of patie nts with total VA. On the other hand, AGA were positive in 0% (zero of two) of patients with Marsh I and in 30% (three of 10) of patients with Marsh I I, whereas EMA were positive in 0% (zero of two) of patients with Marsh I a nd in 40% (four of 10) of patients with Marsh II (Marsh I-IIIa vs Marsh III b-c, p = < 0.005 in overall AGA-positive patients and p = < 0.0001 in EMA-p ositive patients). At the same time the overall AGA in silent form were pos itive in 60% (three of five) of patients with partial VA, in 66.66% (four o f six) of patients with subtotal VA, and in 77.77% (seven of nine) of patie nts with total VA, whereas EMA were positive in 80% (four of five) of patie nts with partial VA, in 83.33% (five of six) of patients with subtotal VA, and in 88.88% (eight of nine) of patients with total VA. On the other hand, overall AGA were positive in 0% of patients with both Marsh I (zero of one ) and Marsh II (zero of two), as well as EMA were positive in 0% with both Marsh I (zero of one) and Marsh II (zero of two) (Marsh I-IIIa vs Marsh III b-c, p = < 0.001 in overall AGA-positive patients and p = < 0.007 in EMA-po sitive patients). CONCLUSIONS: At this time small bowel biopsy seems to be the only correct p rocedure to diagnose a case of suspected celiac disease, especially in pati ents with mild symptoms or suspected for celiac disease, because they belon g to highrisk groups. (C) 2001 by Am. Cell. of Gastroenterology.